Remove semi colon on all transforms

[ahuang11]

I think at some point, I encountered some Lumen AI syntax error related to ; or comments

WITH gold_medals AS (
    SELECT
        "athlete_full_name",
        COUNT(*) AS gold_count
    FROM read_csv('olympic_medals.csv')
    WHERE "medal_type" = 'GOLD'
      AND "slug_game" LIKE '%-2022'  -- Assuming Winter Olympics are identified by the year
    GROUP BY "athlete_full_name"
)
SELECT
    "athlete_full_name",
    gold_count
FROM gold_medals
ORDER BY gold_count DESC
LIMIT 1

Not sure if this will fix, but I think it should be applied to all transforms anyways.

Ideally, these transforms would be applied to the generated SQL for visibility, and not after under the hood in pipeline (okay applied)

[codecov]

Codecov Report

Attention: Patch coverage is 33.33333% with 4 lines in your changes missing coverage. Please review.

Project coverage is 57.69%. Comparing base (60ea02f) to head (1b746ee).

Files with missing lines	Patch %	Lines
lumen/ai/agents.py	0.00%	4 Missing ⚠️

Additional details and impacted files

@@            Coverage Diff             @@
##             main    #1022      +/-   ##
==========================================
- Coverage   57.70%   57.69%   -0.01%     
==========================================
  Files         109      109              
  Lines       14197    14199       +2     
==========================================
  Hits         8192     8192              
- Misses       6005     6007       +2     

☔ View full report in Codecov by Sentry.
📢 Have feedback on the report? Share it here.

[philippjfr]

Confused, why iterate over a loop when there is only a single item? Also why bake the transform into the query?

[ahuang11]

Was just keeping the original structure, in case there were more transforms. I'm fine with dropping the loop.

Applying it on the original query so that it shows up in the SQL code editor, in case there's a bug in the SQLTransform output.

[philippjfr]

Baking it into the query means it falls outside of the programmable interface, e.g. if we want to toggle the limit on and off. So I'm -1 on that part.

[ahuang11]

On the flip side, if the transform (limit) isn't visible to the user, the user might think it's an issue with the data (unaware there's a limit to 1 million rows; the "Full data" checkbox is pretty tiny and I didn't notice it initially because I was looking at the query):

if we want to toggle the limit on and off

I don't follow; can't it still be toggled on and off (e.g. SELECT * FROM read_parquet('num.parquet') AS num_table LIMIT 1000000, changes into SELECT * FROM read_parquet('num.parquet') AS num_table?

By the way, clicking "Full data" checkbox doesn't work as expected: